The 20 Art Styles Challenge asks you to redraw one character twenty times, each time in a different, recognisable art style: anime, Disney-era animation, Cubism, ukiyo-e, pixel art, and so on. BrushWarriors is the version we run as a community prompt, with a fixed list of twenty styles so results are comparable. You complete it by picking one character, working through the list at one to two styles per sitting, and keeping every piece at the same canvas size so the set reads as a grid when posted. The point is not twenty finished paintings. It is forcing yourself to copy the rules of styles you would never choose, which exposes habits you did not know you had.

Which 20 styles are in the BrushWarriors list?

The list mixes art-historical movements, animation house styles and digital-native genres. Historical styles come with documented rules, which is what makes them useful practice. Tate’s glossary entries on Impressionism, Cubism and Pop art are short enough to read before each attempt.

Ukiyo-e in particular rewards research before drawing. The Metropolitan Museum’s essay on woodblock prints in the ukiyo-e style explains why the colour sits in flat registered blocks, which tells you exactly how to build your layers.

How long does the challenge take?

Budget 60 to 90 minutes per style for a half-body character at 2000 by 2000 pixels. That puts the full set at 20 to 30 hours, or about a month at one piece per day. Photorealism and watercolour will run long. Pixel art and vector flat will run short, and that imbalance is part of the lesson.

Set a timer and stop when it goes off. The challenge measures whether you can hit a style’s rules quickly, not whether you can polish. Unfinished pieces that nail the style beat finished pieces that drift back into your default.

What software do you need?

Any raster painting app with layers, blend modes and a custom brush engine will cover all twenty styles. Pixel art and vector flat are the exceptions, and both have free tools.

• Krita is free and open source, runs on Windows, macOS and Linux, and its manual covers brush engines, the wrap-around pattern mode, and a pixel-art brush preset.
• Procreate is a one-time purchase on iPad and the fastest way to test the watercolour and ink-wash styles because of its wet-mix brush behaviour. The Procreate Handbook documents brush studio settings. Procreate stated publicly in 2024 that it will not add generative AI, which matters if you want your entries to be unambiguously hand-made.
• Photoshop remains the standard where client files go, and Adobe’s painting tools documentation is where to learn the Mixer Brush for the Impressionism entry.
• Blender’s Grease Pencil is a free route to the low-poly and comic entries in actual 3D, and the Grease Pencil manual covers line art modifiers that mimic inked outlines.

A pressure-sensitive tablet matters more than the app. Entry-level tablets without screens cost roughly the same as a mid-range game and are enough for the whole challenge. Pen displays cost several times more and mainly speed up the photorealism and watercolour entries.

Where does AI fit, and where does it not?

Generative tools can produce twenty styles of one character in minutes, and that output is worthless for this challenge. The skill being trained is observing a style’s rules and executing them by hand. Using AI for the entries skips the only step that has value.

AI is useful for reference, not production. Generating a mood board of Art Nouveau poster layouts to study before drawing is a legitimate use. Submitting the generation is not. There is also a legal dimension. The US Copyright Office’s Copyright and Artificial Intelligence reports, including the January 2025 Part 2 report on copyrightability, state that prompts alone do not make a person the author of AI output. Hand-drawn entries are yours. Generated ones may not be registrable.

What does a production studio actually learn from this?

We run a shorter version of this exercise on every animated explainer we take on. Before a client approves a style, we produce three to five style frames of the same scene in different visual languages. That round typically takes one to two weeks and is where projects are won or lost, because a client who cannot see the difference between flat vector and textured picture-book will ask for both at once.

The failure mode we see most often from illustrators who have not done a style challenge is drift. They begin a project in the approved style and by shot twelve are back in their default. Twenty forced style switches build the muscle that prevents this, and we hire from portfolios that show it.

The other lesson is cost. Styles with visible texture, painted shading or photoreal rendering take two to four times longer per frame than flat vector, and that multiplies across every second of animation. When a client’s budget is fixed, the style choice is the budget choice. Artists who have personally felt the time difference between entry 19 and entry 20 in the table above can explain that trade-off to a client without guessing.

How should you post the results?

Post the set as a single grid image first, then individual entries. Keep every canvas the same size and the character in the same pose and framing so viewers compare style, not composition. Label each tile with the style name. The grid format is why the challenge spreads: it reads instantly on a phone screen and invites people to pick a favourite.

Hold back the two or three entries you dislike most and post them with a note on what went wrong. Those posts reliably get more replies than the polished ones, and the feedback is more specific.

Pick a character you can draw from memory, set a 90-minute timer, and start with entry 11, pixel art. Its constraints are so tight that there is no room to drift into habit, which makes it the clearest demonstration of what the other nineteen are for.